freebsd-skq/share/man/man9/namei.9
John Baldwin 1d800a676e Fix spelling of EACCES.
PR:		docs/144112
Submitted by:	Alexander Best
MFC after:	3 days
2010-02-24 13:13:29 +00:00

379 lines
9.1 KiB
Groff

.\"
.\" Copyright (c) 1998, 1999 Eivind Eklund
.\" Copyright (c) 2003 Hiten M. Pandya
.\" Copyright (c) 2005 Robert N. M. Watson
.\"
.\" All rights reserved.
.\"
.\" This program is free software.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\" notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\" notice, this list of conditions and the following disclaimer in the
.\" documentation and/or other materials provided with the distribution.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE DEVELOPERS ``AS IS'' AND ANY EXPRESS OR
.\" IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
.\" OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
.\" IN NO EVENT SHALL THE DEVELOPERS BE LIABLE FOR ANY DIRECT, INDIRECT,
.\" INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
.\" NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
.\" DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
.\" THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
.\" (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
.\" THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
.\"
.\"
.\" If you integrate this manpage in another OS, I'd appreciate a note
.\" - eivind@FreeBSD.org
.\"
.\" $FreeBSD$
.\"
.Dd September 21, 2005
.Os
.Dt NAMEI 9
.Sh NAME
.Nm namei ,
.Nm NDINIT ,
.Nm NDFREE ,
.Nm NDHASGIANT
.Nd pathname translation and lookup operations
.Sh SYNOPSIS
.In sys/param.h
.In sys/fcntl.h
.In sys/namei.h
.Ft int
.Fn namei "struct nameidata *ndp"
.Ft void
.Fo NDINIT
.Fa "struct nameidata *ndp" "u_long op" "u_long flags"
.Fa "enum uio_seg segflg" "const char *namep" "struct thread *td"
.Fc
.Ft void
.Fn NDFREE "struct nameidata *ndp" "const uint flags"
.Ft int
.Fn NDHASGIANT "struct nameidata *ndp"
.Sh DESCRIPTION
The
.Nm
facility allows the client to perform pathname translation and lookup
operations.
The
.Nm
functions will increment the reference count for the vnode in question.
The reference count has to be decremented after use of the vnode, by
using either
.Xr vrele 9
or
.Xr vput 9 ,
depending on whether the
.Dv LOCKLEAF
flag was specified or not.
If the
.Va Giant
lock is required,
.Nm
will acquire it if the caller indicates it is
.Dv MPSAFE ,
in which case the caller must later release
.Va Giant
based on the results of
.Fn NDHASGIANT .
.Pp
The
.Fn NDINIT
function is used to initialize
.Nm
components.
It takes the following arguments:
.Bl -tag -width ".Fa segflg"
.It Fa ndp
The
.Vt "struct nameidata"
to initialize.
.It Fa op
The operation which
.Fn namei
will perform.
The following operations are valid:
.Dv LOOKUP , CREATE , DELETE ,
and
.Dv RENAME .
The latter three are just setup for those
effects; just calling
.Fn namei
will not result in
.Fn VOP_RENAME
being called.
.It Fa flags
Operation flags.
Several of these can be effective at the same time.
.It Fa segflg
UIO segment indicator.
This indicates if the name of the object is in userspace
.Pq Dv UIO_USERSPACE
or in the kernel address space
.Pq Dv UIO_SYSSPACE .
.It Fa namep
Pointer to the component's pathname buffer
(the file or directory name that will be looked up).
.It Fa td
The thread context to use for
.Nm
operations and locks.
.El
.Sh NAMEI OPERATION FLAGS
The
.Fn namei
function takes the following set of
.Dq "operation flags"
that influence its operation:
.Bl -tag -width ".Dv WANTPARENT"
.It Dv LOCKLEAF
Lock vnode on return.
This is a full lock of the vnode; the
.Xr VOP_UNLOCK 9
should be used
to release the lock (or
.Xr vput 9
which is equivalent to calling
.Xr VOP_UNLOCK 9
followed by
.Xr vrele 9 ,
all in one).
.It Dv LOCKPARENT
This flag lets the
.Fn namei
function return the parent (directory) vnode,
.Va ni_dvp
in locked state, unless it is identical to
.Va ni_vp ,
in which case
.Va ni_dvp
is not locked per se (but may be locked due to
.Dv LOCKLEAF ) .
If a lock is enforced, it should be released using
.Xr vput 9
or
.Xr VOP_UNLOCK 9
and
.Xr vrele 9 .
.It Dv WANTPARENT
This flag allows the
.Fn namei
function to return the parent (directory) vnode in an unlocked state.
The parent vnode must be released separately by using
.Xr vrele 9 .
.It Dv MPSAFE
With this flag set,
.Fn namei
will conditionally acquire
.Va Giant
if it is required by a traversed file system.
MPSAFE callers should pass the results of
.Fn NDHASGIANT
to
.Xr VFS_UNLOCK_GIANT
in order to conditionally release
.Va Giant
if necessary.
.It Dv NOCACHE
Avoid
.Fn namei
creating this entry in the namecache if it is not
already present.
Normally,
.Fn namei
will add entries to the name cache
if they are not already there.
.It Dv FOLLOW
With this flag,
.Fn namei
will follow the symbolic link if the last part
of the path supplied is a symbolic link (i.e., it will return a vnode
for whatever the link points at, instead for the link itself).
.It Dv NOOBJ
Do not call
.Fn vfs_object_create
for the returned vnode, even though it meets required criteria for VM support.
.It Dv NOFOLLOW
Do not follow symbolic links (pseudo).
This flag is not looked for by the actual code, which looks for
.Dv FOLLOW .
.Dv NOFOLLOW
is used to indicate to the source code reader that symlinks
are intentionally not followed.
.It Dv SAVENAME
Do not free the pathname buffer at the end of the
.Fn namei
invocation; instead, free it later in
.Fn NDFREE
so that the caller may access the pathname buffer.
See below for details.
.It Dv SAVESTART
Retain an additional reference to the parent directory; do not free
the pathname buffer.
See below for details.
.El
.Sh ALLOCATED ELEMENTS
The
.Vt nameidata
structure is composed of the following fields:
.Bl -tag -width ".Va ni_cnd.cn_pnbuf"
.It Va ni_startdir
In the normal case, this is either the current directory or the root.
It is the current directory if the name passed in does not start with
.Ql /
and we have not gone through any symlinks with an absolute path, and
the root otherwise.
.Pp
In this case, it is only used by
.Fn lookup ,
and should not be
considered valid after a call to
.Fn namei .
If
.Dv SAVESTART
is set, this is set to the same as
.Va ni_dvp ,
with an extra
.Xr vref 9 .
To block
.Fn NDFREE
from releasing
.Va ni_startdir ,
the
.Dv NDF_NO_STARTDIR_RELE
can be set.
.It Va ni_dvp
Vnode pointer to directory of the object on which lookup is performed.
This is available on successful return if
.Dv LOCKPARENT
or
.Dv WANTPARENT
is set.
It is locked if
.Dv LOCKPARENT
is set.
Freeing this in
.Fn NDFREE
can be inhibited by
.Dv NDF_NO_DVP_RELE , NDF_NO_DVP_PUT ,
or
.Dv NDF_NO_DVP_UNLOCK
(with the obvious effects).
.It Va ni_vp
Vnode pointer to the resulting object,
.Dv NULL
otherwise.
The
.Va v_usecount
field of this vnode is incremented.
If
.Dv LOCKLEAF
is set, it is also locked.
.Pp
Freeing this in
.Fn NDFREE
can be inhibited by
.Dv NDF_NO_VP_RELE , NDF_NO_VP_PUT ,
or
.Dv NDF_NO_VP_UNLOCK
(with the obvious effects).
.It Va ni_cnd.cn_pnbuf
The pathname buffer contains the location of the file or directory
that will be used by the
.Nm
operations.
It is managed by the
.Xr uma 9
zone allocation interface.
If the
.Dv SAVESTART
or
.Dv SAVENAME
flag is set, then the pathname buffer is available
after calling the
.Fn namei
function.
.Pp
To only deallocate resources used by the pathname buffer,
.Va ni_cnd.cn_pnbuf ,
then
.Dv NDF_ONLY_PNBUF
flag can be passed to the
.Fn NDFREE
function.
To keep the pathname buffer intact,
the
.Dv NDF_NO_FREE_PNBUF
flag can be passed to the
.Fn NDFREE
function.
.El
.Sh RETURN VALUES
If successful,
.Fn namei
will return 0, otherwise it will return an error.
.Sh ERRORS
Errors which
.Fn namei
may return:
.Bl -tag -width Er
.It Bq Er ENOTDIR
A component of the specified pathname is not a directory when a directory is
expected.
.It Bq Er ENAMETOOLONG
A component of a pathname exceeded 255 characters,
or an entire pathname exceeded 1023 characters.
.It Bq Er ENOENT
A component of the specified pathname does not exist,
or the pathname is an empty string.
.It Bq Er EACCES
An attempt is made to access a file in a way forbidden by its file access
permissions.
.It Bq Er ELOOP
Too many symbolic links were encountered in translating the pathname.
.It Bq Er EISDIR
An attempt is made to open a directory with write mode specified.
.It Bq Er EROFS
An attempt is made to modify a file or directory on a read-only file system.
.El
.Sh FILES
.Bl -tag
.It Pa src/sys/kern/vfs_lookup.c
.El
.Sh SEE ALSO
.Xr uio 9 ,
.Xr uma 9 ,
.Xr VFS 9 ,
.Xr VFS_UNLOCK_GIANT 9 ,
.Xr vnode 9 ,
.Xr vput 9 ,
.Xr vref 9
.Sh AUTHORS
.An -nosplit
This manual page was written by
.An Eivind Eklund Aq eivind@FreeBSD.org
and later significantly revised by
.An Hiten M. Pandya Aq hmp@FreeBSD.org .
.Sh BUGS
The
.Dv LOCKPARENT
flag does not always result in the parent vnode being locked.
This results in complications when the
.Dv LOCKPARENT
is used.
In order to solve this for the cases where both
.Dv LOCKPARENT
and
.Dv LOCKLEAF
are used, it is necessary to resort to recursive locking.
.Pp
Non-MPSAFE file systems exist, requiring callers to conditionally unlock
.Va Giant .